We calculated the expected predictive power of each PGI (that might potentially be included in the Repository) using the following formula from Daetwyler et al.75: E(R2)=(hSNP2)2hSNP2+MN, where hSNP2 is the phenotype’s SNP heritability, M is the effective number of independent SNPs which we assume to be equal to 60,0009, and N is the GWAS sample size for the phenotype.
